JP7063680B2 - Image processing equipment, foreign matter inspection equipment, image processing method, and foreign matter inspection method - Google Patents

Description

The present invention relates to an image processing apparatus, a foreign matter inspection apparatus, an image processing method, and a foreign matter inspection method.

Non-aqueous electrolyte secondary batteries such as lithium ion secondary batteries are widely used as batteries for personal computers, mobile phones, mobile information terminals and the like. In particular, the lithium ion secondary battery is attracting attention as a battery that reduces CO ₂ emissions and contributes to energy saving as compared with the conventional secondary battery.

Conventionally, the development of a separator winder in which a separator for a non-aqueous electrolytic solution secondary battery is wound around a core has been promoted. At the same time, a foreign matter inspection for detecting a foreign matter adhering to the separator winding body is being studied.

As an example of the inspection applicable to the foreign matter inspection, the technique disclosed in Patent Document 1 can be mentioned. In the technique disclosed in Patent Document 1, defects are found by integrating the density values of each pixel from the image data of the inspection target and calculating the integrated value, and comparing the integrated value with a predetermined threshold value. To detect.

Japanese Unexamined Patent Publication No. 2005-265467 (published on September 29, 2005)

In the technique disclosed in Patent Document 1, the presence or absence of a foreign substance that is a defect can be detected, and the planar size of the foreign substance can be estimated based on the range corresponding to the foreign substance. The problem arises that it is difficult to estimate the three-dimensional size.

One aspect of the present invention is to realize an image processing apparatus, a foreign matter inspection apparatus, an image processing method, and a foreign matter inspection method capable of estimating a three-dimensional size of a foreign matter.

In order to solve the above-mentioned problems, the image processing apparatus according to one aspect of the present invention constitutes the above-mentioned image in an image processing apparatus that performs image processing on an image of an electromagnetic wave transmitted through an inspection object containing a foreign substance. A storage unit that stores the first pixel value corresponding to each of a large number of pixels, and a background corresponding to each of the first pixel values and the first pixel value after attenuation by the inspection object containing no foreign matter in each of the pixels. It includes a pixel value calculation unit that calculates the second pixel value of each pixel, which is proportional to the difference from the value, and a pixel value integration unit that integrates each second pixel value of the pixel group belonging to a specific continuous region. ..

In order to solve the above-mentioned problems, the image processing method according to one aspect of the present invention constitutes the above-mentioned image in an image processing method for performing image processing on an image of an electromagnetic wave transmitted through an inspection object containing a foreign substance. A storage process for storing the first pixel value corresponding to each of a large number of pixels, and a background corresponding to each of the first pixel values and the first pixel value after attenuation by an inspection object containing no foreign matter in each of the pixels. It includes a pixel value calculation step of calculating the second pixel value of each pixel, which is proportional to the difference from the value, and a pixel value integration step of integrating each second pixel value of a pixel group belonging to a specific continuous region. ..

In particular, in the image processing apparatus and image processing method according to one aspect of the present invention, the second pixel value of each pixel is the background of the difference between the first pixel value and the background value in each pixel. The ratio to the value.

According to the above configuration, the second pixel value is a value corresponding to the thickness of the portion corresponding to the pixel in the foreign matter. Then, the value obtained by integrating the second pixel values of each of the plurality of pixels belonging to the specific continuous region becomes a value corresponding to the volume of the portion of the foreign substance corresponding to the plurality of pixels. Therefore, since the volume of a part or all of the foreign matter can be estimated, the presence or absence of the foreign matter having a certain volume or more can be determined. That is, according to the above configuration, it is possible to estimate the three-dimensional size of the foreign matter.

In the image processing apparatus according to one aspect of the present invention, the pixel value integrating unit uses each of the large number of pixels as the pixel of interest and integrates the second pixel value of each pixel group belonging to the continuous region including the pixel of interest. do.

In the image processing method according to one aspect of the present invention, in the pixel value integration step, each of the large number of pixels is set as a pixel of interest, and each second pixel value of the pixel group belonging to the continuous region including the pixel of interest is set. Accumulate.

According to the above configuration, a continuous region substantially corresponding to the volume of the foreign matter set as the lower limit of detection is set, and the integrated value corresponding to the foreign matter is used as the threshold value to obtain the foreign matter having a volume equal to or larger than the lower limit of detection. It can be detected selectively.

In the image processing apparatus according to one aspect of the present invention, the pixel value integrating unit has a second pixel value indicating a value whose attenuation is larger than a threshold value, and is a second pixel of each of a plurality of pixels forming a continuous region. Pixel values are integrated.

The image processing method according to one aspect of the present invention has a second pixel value indicating a value whose attenuation is larger than a threshold value in the pixel value integration step, and is the first of each of a plurality of pixels forming a continuous region. 2 Pixel values are integrated.

According to the above configuration, a plurality of pixels having a second pixel value indicating a value whose attenuation is larger than the threshold value and forming a continuous region can be regarded as corresponding to one independent foreign substance. .. Therefore, the volume of one independent foreign body can be estimated.

The foreign matter inspection device according to one aspect of the present invention includes the above-mentioned image processing device.

The foreign matter inspection method according to one aspect of the present invention includes the above image processing method.

According to the above configuration, the foreign matter inspection apparatus and the foreign matter inspection method have the same effects as the image processing apparatus and the image processing method, respectively.

According to one aspect of the present invention, it is possible to estimate the three-dimensional size of a foreign substance.

It is a figure which shows the schematic structure of the foreign matter inspection apparatus which concerns on one aspect of this invention. It is a top view of the main surface of an image sensor. It is a top view which shows the setting example of a plurality of reference pixels. It is a flowchart which shows the flow of image processing by an image processing apparatus. Each first pixel value stored by the storage unit is shown in association with each of the large number of pixels shown in FIG. Each second pixel value calculated by the pixel value calculation unit based on each first pixel value shown in FIG. 5 and the background value set for each pixel corresponds to each of a large number of pixels shown in FIG. It is attached and shown. Based on each second pixel value shown in FIG. 6, each integrated value integrated by the pixel value integrating unit is shown in association with each of the large number of pixels shown in FIG. It is a perspective view which shows the quadrangular pyramid which consists of the center of the electromagnetic wave generation part in the electromagnetic wave generation source, and the four corners of a certain pixel. It is a figure which shows the schematic structure of the 1st modification of the foreign matter inspection apparatus shown in FIG. 1. It is a figure which shows the schematic structure of the 2nd modification of the foreign matter inspection apparatus shown in FIG. It is a figure explaining the rotation of an inspection object by a rotation mechanism. It is a table which shows the result of having verified the effectiveness of the pixel value integrating part in the foreign matter inspection apparatus which used X-ray as an electromagnetic wave.

FIG. 1 is a diagram showing a schematic configuration of a foreign matter inspection device 100 according to an aspect of the present invention. FIG. 2 is a plan view of the main surface 31 of the image sensor 3.

The foreign matter inspection device 100 detects the foreign matter 13 adhering to the inspection object 1. The inspection object 1 is a separator wound body in which a separator 12 for a non-aqueous electrolytic solution secondary battery is wound around a core 11. However, the inspection object 1 may be a film other than the separator 12 for a non-aqueous electrolytic solution secondary battery wound around the core 11, or the film is wound around the core 11. It does not have to be. The foreign matter inspection device 100 includes an electromagnetic wave generation source 2, an image sensor 3, and an image processing device 4.

The electromagnetic wave generation source 2 irradiates one of the side surfaces of the inspection object 1 with the electromagnetic wave 21. The electromagnetic wave 21 passes through the inspection object 1 and the foreign matter 13 and exits from the other side of the inspection object 1. Each side surface of the inspection object 1 is a surface formed by each of the widthwise end portions of the separator 12 for a non-aqueous electrolytic solution secondary battery. The width direction of the separator 12 for a non-aqueous electrolyte secondary battery is the direction along the surface of the separator 12 for a non-aqueous electrolyte secondary battery, and is used in the manufacturing process of the separator 12 for a non-aqueous electrolyte secondary battery. The direction is perpendicular to the direction in which the separator 12 for the non-aqueous electrolytic solution secondary battery is conveyed. Here, as the electromagnetic wave 21, X-rays radially emitted from the electromagnetic wave generation source 2 are applied.

The image sensor 3 has a main surface 31. The main surface 31 is provided with a large number of pixels 32 that receive the electromagnetic wave 21 transmitted through the inspection object 1 and the foreign matter 13 and constitute an image of the electromagnetic wave 21. For convenience of illustration, FIG. 2 shows an excerpt of 225 pixels 32 arranged in a matrix of 15 rows and 15 columns from a large number of pixels 32. Here, an X-ray image sensor is applied as the image sensor 3.

It can be said that the image composed of a large number of pixels 32 is an image obtained by photographing the inspection object 1 forming the background in the inspection image and the foreign matter 13 having contrast with respect to the background.

In the present embodiment, the foreign matter 13 will be described as being a material that attenuates the transmitted X-rays, such as a metallic foreign matter, more than the inspection target 1. In this case, the image composed of a large number of pixels 32 is an image of the electromagnetic wave 21 that has passed through the inspection object 1 including the foreign matter 13 and is more attenuated by the foreign matter 13 attached to the inspection object 1 than the inspection object 1. It can also be said that.

On the contrary, the foreign matter 13 may be made of a material that attenuates the transmitted X-rays, such as a cavity defect (air bubble) formed in the inspection object 1, to be smaller than that of the inspection object 1. In this case, the image composed of a large number of pixels 32 is an image of the electromagnetic wave 21 that has passed through the inspection object 1 including the foreign matter 13 and is attenuated smaller by the foreign matter 13 attached to the inspection object 1 than the inspection object 1. It can also be said that.

In any case, the image composed of a large number of pixels 32 is an image obtained by photographing the inspection object 1 forming the background in the inspection image and the foreign matter 13 having contrast with respect to the background.

The image processing device 4 performs image processing on an image composed of a large number of pixels 32. The image processing device 4 includes a storage unit 41, a pixel value calculation unit 42, a background value setting unit 43, and a pixel value integration unit 44.

The storage unit 41 is composed of, for example, a well-known storage medium. The storage unit 41 stores pixel values corresponding to each of a large number of pixels 32 that constitute the above image for which the image processing device 4 performs image processing. Hereinafter, the pixel value stored in the storage unit 41 is referred to as a first pixel value. In the present specification, the first pixel value and the background value described later are assumed to be larger as the corresponding image is brighter, but the magnitude relationship with respect to the brightness of the image may be reversed.

FIG. 3 is a plan view showing a setting example of a plurality of reference pixels 34. In FIG. 3, the plurality of reference pixels 34 are the plurality of pixels 32 shown in shading. As will be described later, the reference pixel 34 is a pixel 32 necessary for setting the background value of the corresponding pixel of interest 33.

The pixel value calculation unit 42 is configured by, for example, a CPU (Central Processing Unit) or hardware logic. Hereinafter, the pixel of interest 33 shown in FIG. 3 will be described as an example. As will be described later, the pixel of interest 33 is a pixel 32 whose background value is set by the corresponding plurality of reference pixels 34. The pixel value calculation unit 42 uses each of the large number of pixels 32 as the pixel of interest 33, and calculates the second pixel value of the pixel of interest 33 based on the first pixel value stored in the storage unit 42. In other words, the pixel value calculation unit 42 calculates the second pixel value of each pixel 32.

Here, with respect to each of the pixel of interest 33 and each of the above pixels 32, the second pixel value is the ratio of the difference between the first pixel value and the background value to the background value. It can be said that the background value corresponds to the amount of attenuation of the electromagnetic wave 21 by the inspection object 1. That is, the second pixel value is "(first pixel value-background value) / background value" with respect to each of the pixel of interest 33 and each of the above pixels 32. The calculation result of the second pixel value is also called a relative density because the ratio of the difference to the background value is obtained.

The background value setting unit 43 is configured by, for example, a CPU or hardware logic. The background value setting unit 43 has a background value based on the first pixel value of the reference pixel 34, which is located in the vicinity of the pixel of interest 33 and has a predetermined positional relationship with respect to the pixel of interest 33. A value to be regarded as is determined, and that value is set as a background value of the pixel of interest 33. Each pixel of interest 33 referred to here corresponds to each of the pixels 32.

The pixel value integrating unit 44 integrates each second pixel value of the pixel group belonging to a specific continuous region.

FIG. 4 is a flowchart showing the flow of image processing by the image processing device 4.

When a large number of pixels 32 form an image of the electromagnetic wave 21, the storage unit 41 first stores the first pixel value corresponding to each of the large number of pixels 32. This process corresponds to step S1 in FIG. FIG. 5 shows each first pixel value stored by the storage unit 41 in association with each of the large number of pixels 32 shown in FIG. In FIG. 5, among the above-mentioned 225 pixels 32, the central 5 × 5 = 25 pixels 32 constitute an image of the electromagnetic wave 21 attenuated by the inspection object 1 and the foreign matter 13. An example is shown in which the pixel 32 constitutes an image of the electromagnetic wave 21 attenuated by the inspection object 1. The 25 pixels 32 in the center correspond to the pixel group 37. In FIG. 5, the definitions of A1 and A2 are as follows.

A1: First pixel value corresponding to the attenuation amount of the electromagnetic wave 21 due to the inspection object 1 A2: First pixel value corresponding to the attenuation amount of the electromagnetic wave 21 due to the inspection object 1 and the foreign matter 13 Next, the background value setting unit 43 , A background value is set for each pixel 32. This process corresponds to step S2 in FIG. By setting the background value by the background value setting unit 43, the foreign matter inspection device 100 has the following effects.

That is, according to the above configuration, the background value is set based on the first pixel value of the reference pixel 34 located in the vicinity of the attention pixel 33. As a result, for example, in a transmission inspection using X-rays, a relatively high inspection target 1 having a slight variation in transmittance and a foreign substance 13 contained in the inspection target 1 having a relatively low transmittance are detected. In this case, it is possible to obtain a second pixel value in which the influence of the variation is reduced. Then, by detecting the foreign matter 13 with reference to the second pixel value, high-precision foreign matter inspection becomes possible.

That is, according to the above configuration, it is possible to cancel the distribution of the transmittance in the inspection object 1. For example, even if the contour line in the slope of the transmittance with respect to the position in the inspection object 1 corresponding to the distribution is not a straight line, if the curvature of the contour line is low, the background value is set almost accurately, that is, with a slight error. As a result, the correct second pixel value can be calculated.

By the way, in the present embodiment, the electromagnetic wave generation source 2 generates X-rays by hitting an electron on a metal plate, but since the place where the electron hits is heated, the place where the electron hits the metal plate is changed. The metal plate is vibrated. As a result, the electromagnetic wave generation source 2 generates the X-rays in a planar shape instead of a dot shape. As a result, when the electromagnetic wave generation source 2 irradiates the X-ray as the electromagnetic wave 21, the distribution of the brightness of the image of the electromagnetic wave 21 may vary every time the image is taken. Further, due to the noise generated in the image of the electromagnetic wave 21, the distribution of the brightness of the image of the electromagnetic wave 21 may vary every time the image is taken. Even in such a case, according to the above configuration, it is possible to obtain a second pixel value in which the influence of the above variation is reduced.

By setting the reference pixel 34 while avoiding the pixel 32 having an extremely large first pixel value and the pixel 32 having an extremely small first pixel value, it is possible to prevent the background value from being significantly deviated from the desired value. As a result, it is possible to obtain a second pixel value in which the influence of the above variation is reduced.

At least one reference pixel 34 may be set, but it is preferable that the number of reference pixels 34 is a plurality as shown in FIG. As a result, the number of first pixel values that can be used to determine the value to be regarded as the background value increases, so that the options for setting the background value can be expanded. The options will be described below.

As shown in FIG. 3, the plurality of reference pixels 34 are preferably a plurality of pixels 32 arranged so as to surround the corresponding pixel of interest 33.

According to the above configuration, when unsuitable pixels 35 having a first pixel value that is not suitable as a value for setting a background value are concentrated in a certain reference pixel 34 and its vicinity, a plurality of references are made. The ratio of the unsuitable pixel 35 to the pixel 34 can be reduced. Therefore, it is possible to reduce the possibility that the first pixel value of the unsuitable pixel 35 adversely affects the setting of the background value.

That is, in FIG. 3, assuming that each of the shaded pixels 32 is an unsuitable pixel 35, a plurality of unsuitable pixels 35 are concentrated in the upper left of the figure. According to the above configuration, there are reference pixels 34 that correspond to the unsuitable pixels 35, but there are many reference pixels 34 that do not correspond to the unsuitable pixels 35. In this way, the ratio of the unsuitable pixel 35 to the plurality of reference pixels 34 can be reduced.

As an example of the unsuitable pixel 35, the pixel 32 constituting the image of the electromagnetic wave 21 transmitted through the tape or the label 14 attached to the inspection object 1 can be mentioned. The first pixel value obtained from these pixels 32 is an image of the electromagnetic wave 21 transmitted through the inspection object 1 itself, and is compared with the first pixel value obtained from the pixels 32 constituting the shadow-free image. Therefore, it tends to be an extremely small value.

As shown in FIG. 3, the plurality of reference pixels 34 are preferably a plurality of pixels 32 arranged symmetrically in the row direction and the column direction with respect to the corresponding pixel of interest 33. The row direction and the column direction correspond to the "row direction" and the "column direction" in FIG. 3, respectively.

Further, it can be said that the plurality of reference pixels 34 shown in FIG. 3 have the pixel group 36 arranged rotationally symmetrically four times around the pixel of interest 33. As described above, the plurality of reference pixels 34 are preferably a plurality of pixels 32 arranged symmetrically by n (n is an integer of 2 or more) around the corresponding pixel of interest 33.

According to the above configuration, the ratio of the unsuitable pixel 35 to the plurality of reference pixels 34 is reduced even when the unsuitable pixel 35 is concentrated in any direction with respect to the attention pixel 33. be able to. Therefore, it is possible to reduce the possibility that the first pixel value of the unsuitable pixel 35 adversely affects the setting of the background value.

That is, in FIG. 3, a plurality of unsuitable pixels 35 are concentrated in the direction from the attention pixel 33 toward the upper left in the figure. According to the above configuration, there are reference pixels 34 that correspond to the unsuitable pixels 35, but there are many reference pixels 34 that do not correspond to the unsuitable pixels 35. In this way, the ratio of the unsuitable pixel 35 to the plurality of reference pixels 34 can be reduced.

For example, when it is necessary to change the direction in which the inspection object 1 is photographed in the foreign matter inspection device 100, the position where the plurality of unsuitable pixels 35 are concentrated also changes accordingly. Similarly, when the orientation of the inspection object 1 is unavoidably changed due to the difficulty in sufficiently controlling the orientation of the inspection object 1 in the foreign matter inspection apparatus 100, a plurality of unsuitable pixels 35 are concentrated. The position will change. According to the foreign matter inspection device 100, the ratio of the unsuitable pixel 35 to the plurality of reference pixels 34 can be reduced both before and after the change in the direction in which the inspection object 1 is photographed. Therefore, it can be said that the foreign matter inspection device 100 has little dependence on the direction in which the inspection object 1 is photographed.

The plurality of reference pixels 34 are, but are not limited to, the plurality of pixels 32 arranged so as to surround the pixel of interest 33 without being separated from each other in FIG. That is, at least one reference pixel 34 may be arranged apart from the other reference pixels 34. For example, each reference pixel 34 may be scattered. Further, the plurality of reference pixels 34 may be divided into a plurality of blocks, the blocks may be scattered, and the reference pixels 34 constituting the same block may be adjacent to each other.

The background value setting unit 43 sets the median value of the first pixel value of each of the plurality of reference pixels 34, or the value excluding the maximum value and the minimum value, which is in a predetermined order, as the background value. The median is the average of the p-th largest first pixel value and p + 1st-largest first pixel value when the total number of the first pixel values of each of the plurality of reference pixels 34 is 2p (p is a natural number). It is a value, and when the total number is 2p + 1, it is the first pixel value having the largest p + 1. Further, it is preferable that the predetermined rank among the values excluding the maximum value and the minimum value is a rank in which the original background value can be easily obtained according to the characteristics such as noise. The preferred value for the rank is often as close as possible to half of the total number, but depending on the shape and / or arrangement of the inspection object 1, the rank may be determined using, for example, the concept of the mode. May be preferable. This obtains a brightness histogram for a plurality of reference pixels 34 and determines the predetermined order based on the high frequency of appearance (for example, the order corresponding to the brightness with the highest frequency of appearance is adopted. ) The method.

According to the above configuration, it is possible to easily prevent an extremely large value or an extremely small value from the first pixel values of each of the plurality of reference pixels 34 from being set as the background value. That is, it is possible to prevent the background value from showing an unreasonably extremely bright background or an unreasonably extremely dark background. Further, due to the influence of this, it is possible to easily prevent the background value from greatly deviating from the original value. Therefore, it is possible to set a background value that is more suitable for the captured image.

For example, when the image composed of a reference pixel 34 contains noise, the first pixel value of the reference pixel 34 can be an extremely large value or an extremely small value. It can be said that the foreign matter inspection device 100 reduces the possibility that the first pixel value whose value is disturbed due to such noise is applied as the background value.

The background value setting unit 43 can also be applied in the case of setting a background value corresponding to the amount of attenuation of the electromagnetic wave 21 by the inspection object 1 in the foreign matter inspection in which the electromagnetic wave 21 is reflected by the inspection object 1 and the foreign matter 13. ..

Subsequently, the pixel value calculation unit 42 calculates the second pixel value of each pixel 32. This process corresponds to step S3 in FIG. FIG. 6 shows each second pixel value calculated by the pixel value calculation unit 42 based on each first pixel value shown in FIG. 5 and the background value set for each pixel 32. It is shown in association with each of the large number of pixels 32. In FIG. 6, the definitions of B1 and B2 are as follows.

B1: B1 = (A1-background value of the corresponding pixel 32) / background value of the corresponding pixel 32 B2: B2 = (A2-background value of the corresponding pixel 32) / background value of the corresponding pixel 32 Ideal A1 is the same as the background value of the corresponding pixel 32, and B1 is 0. Each second pixel value is a value corresponding to the thickness of the portion of the foreign matter 13 corresponding to the pixel 32.

Subsequently, the pixel value integrating unit 44 integrates each second pixel value of the pixel group belonging to the specific continuous region. This process corresponds to step S4 in FIG.

FIG. 7 shows each integrated value integrated by the pixel value integrating unit 44 based on each second pixel value shown in FIG. 6 in association with each of the large number of pixels 32 shown in FIG. .. In FIG. 7, each of a large number of pixels 32 is a pixel of interest 33, and a total of nine pixels 32 provided in 3 rows and 3 columns centered on the pixel of interest 33 are continuous regions, and correspond to a pixel group 38 belonging to the continuous region. An example of integrating the second pixel value of each of the nine pixels 32 is shown. In FIG. 7, the definitions of C1 to C7 are as follows.

C1: C1 = 9 * B1
C2: C2 = 8 * B1 + 1 * B2
C3: C3 = 7 * B1 + 2 * B2
C4: C4 = 6 * B1 + 3 * B2
C5: C5 = 5 * B1 + 4 * B2
C6: C6 = 3 * B1 + 6 * B2
C7: C7 = 9 * B2
As described above, the electromagnetic wave generation source 2 generates X-rays in a planar shape instead of a dot shape. As a result, when the electromagnetic wave generation source 2 irradiates the X-ray as the electromagnetic wave 21, the image of the electromagnetic wave 21 is blurred. Therefore, in the image of the foreign matter 13 on the image, a part of the attenuation that should be originally reflected in a certain pixel 32 is reflected in the pixels 32 around the pixel 32.

Further, unlike visible light, it is difficult to narrow the spot diameter or convert X-rays into parallel light by using a lens. Therefore, the image of the foreign matter 13 contained in the image is more likely to be blurred as compared with the case of using visible light.

However, according to the above configuration, since the second pixel value of the pixel group belonging to the specific continuous region is integrated, the size of the continuous region is set to be larger than the size of the foreign matter 13 to be detected. By setting it, the total amount of X-ray attenuation caused by the foreign matter 13 can be calculated. Therefore, the image of the foreign matter 13 contained in the image can be detected. As a result, the risk of omission of detection of the foreign matter 13 can be reduced. Further, even if the image is blurred, the exposure time in image shooting can be shortened as compared with the conventional case by utilizing the advantage that the risk of omission of detection of foreign matter 13 can be reduced. .. It is possible to obtain these effects even if the pixel value integrating unit 44 integrates each first pixel value stored by the storage unit 41 instead of each second pixel value.

To summarize the above, the specific continuous region is at least a region larger than the region including all the pixels 32 constituting the image of the electromagnetic wave 21 attenuated by the inspection object 1 and the foreign matter 13.

Further, it is more preferable that the specific continuous region is a region larger than the region further including all the pixels 32 corresponding to the blurring of the foreign matter 13 in the image. According to this, the image of the foreign matter 13 contained in the above image can be detected more reliably.

The specific continuous region may be appropriately adjusted so as to include at least the pixels corresponding to the blurring of the foreign matter 13 in the image.

Further, the integrated value of the second pixel value of each of the plurality of pixels 32 belonging to the specific continuous region becomes a value corresponding to the volume of the portion of the foreign matter 13 corresponding to the plurality of pixels 32. Therefore, since the volume of a part or all of the foreign matter 13 can be estimated, the presence or absence of the foreign matter 13 having a certain volume or more can be determined. That is, according to the above configuration, it is possible to estimate the three-dimensional size of the foreign matter 13.

The relationship between the amount of attenuation of the electromagnetic wave 21 which is an X-ray and the volume of the foreign matter 13 will be described in detail. The volume of the foreign matter 13 corresponds to the three-dimensional size of the foreign matter 13. Hereinafter, the description will be made assuming that the size of the foreign matter 13 is very small. Specifically, the size of the minute foreign matter 13 is, for example, a size that fits in a sphere having a diameter of 0.05 mm or more and a diameter of 0.3 mm or less.

The transmittance T of the electromagnetic wave 21 is obtained by the following mathematical formula (1). In the mathematical formula (1), a is the absorption coefficient of the substance transmitted by the electromagnetic wave 21, and z is the total length of the electromagnetic wave 21 transmitted through the substance. When the electromagnetic wave 21 is incident on the substance once, z is the thickness of the substance along the transmission direction of the electromagnetic wave 21. As an example of the case where the electromagnetic wave 21 is incident on the substance multiple times, there is a case where there is a cavity inside the substance or the substance is curved.

Further, how dark the image of the electromagnetic wave 21 is due to the transmission of the electromagnetic wave 21 through the foreign matter 13 can be represented by 1-T. This 1-T is called a shielding rate.

FIG. 8 is a perspective view showing a quadrangular pyramid 23 composed of the center 22 of the generation portion of the electromagnetic wave 21 in the electromagnetic wave generation source 2 and the four corners of a certain pixel 32. The calculation for integrating the shielding rate for the foreign matter 13 in the quadrangular pyramid 23 is shown in the following mathematical formula (2). The calculation result of the mathematical formula (2) is the attenuation rate of the intensity of the electromagnetic wave 21 caused by the foreign matter 13 in the pixel 32. In the formula (2), the x direction corresponding to x and the y direction corresponding to y are both perpendicular to the thickness direction of the foreign matter 13 and perpendicular to each other.

Here, when the size of the foreign matter 13 is small, the cross-sectional dimension of the quadrangular pyramid 23 substantially parallel to the surface of the pixel 32 is substantially constant around the foreign matter 13. From this, the quadrangular pyramid 23 can be approximated to a rectangular parallelepiped around the foreign body 13. Further, when z is substantially equal to 0, the following mathematical formula (3) is established.

As a result, the attenuation rate of the intensity of the electromagnetic wave 21 caused by the foreign matter 13 in the pixel 32 can be approximated by the following mathematical formula (4).

Then, the calculation result of the mathematical formula (4) corresponds to the volume of the portion where the rectangular parallelepiped and the foreign matter 13 overlap. From the above, it can be seen that there is a correlation between the amount of attenuation of the electromagnetic wave 21 which is an X-ray and the volume of the foreign matter 13. That is, for each of the large number of pixels 32, the volume of the foreign matter 13 that overlaps with the corresponding rectangular parallelepiped can be obtained based on the above-mentioned relative density.

From here, the mechanism for converting the relative concentration (B1 and B2 defined above) or the integrated value thereof (C1 to C7 defined above) into the volume of the foreign substance 13 will be supplemented. Let L be the first pixel value (ideally equal to the background value) corresponding to the amount of attenuation of the electromagnetic wave 21 by the inspection object 1. The LT obtained by multiplying L by T due to the foreign matter 13 is the actual first pixel value. L is a value different depending on the position in the electromagnetic wave 21. In other words, L is a function L (x, y) of x, y. Since the second pixel value is represented by (LT-L) / L = (T-1), this is multiplied by -1 and integrated for x and y, that is, the mathematical formula (2) is the mathematical formula (4). ) Approximately. By dividing the value obtained by the mathematical formula (2) by the proportional coefficient a, the volume of the foreign matter 13 can be obtained. When the integrated value of the difference between the first pixel value and the background value is used, (T-1) becomes L (T-1). In this case, L needs to be a constant in order to estimate the volume of the foreign matter 13. In other words, it can be estimated that the inspection object 1 does not contain the foreign matter 13 by using the integrated value of the difference.

The pixel value integrating unit 44 uses each of the large number of pixels 32 as the pixel of interest 33, and integrates each second pixel value of the pixel group belonging to the specific continuous region including the pixel of interest 33.

According to the above configuration, by setting a continuous region substantially corresponding to the volume of the foreign matter 13 set as the lower limit of detection and using the integrated value corresponding to the foreign matter 13 as the threshold value, the foreign matter having a volume equal to or larger than the lower limit of detection is used. 13 can be selectively detected.

It is preferable that the pixel value integrating unit 44 has a second pixel value indicating a value whose attenuation is larger than the threshold value, and integrates the second pixel values of each of the plurality of pixels 32 forming the continuous region. The threshold value is set so that it can be clearly grasped that the electromagnetic wave 21 is attenuated by the foreign matter 13 for each of a large number of pixels 32 without being affected by noise or blurring generated in the image of the electromagnetic wave 21. It is preferable to do so.

A plurality of pixels 32 having a second pixel value indicating a value having an attenuation larger than the threshold value and forming a continuous region can be regarded as corresponding to one independent foreign matter 13. Therefore, the volume of one independent foreign body 13 can be estimated.

Further, the pixel value integration unit 44 may weight at least two second pixel values to be integrated.

For example, when the foreign matter 13 is a sphere, it is expected that the image becomes brighter because the volume of the foreign matter 13 overlapping with the electromagnetic wave 21 is smaller toward the edge from the center of the foreign matter 13. Then, for the purpose of maximizing the value obtained corresponding to the foreign matter 13, as an example of the weighting, the contribution to the integrated value becomes smaller as the second pixel value corresponding to the position closer to the edge of the foreign matter 13 becomes smaller. In addition, it is conceivable to weight each second pixel value at the time of integration.

In this way, the pixel value integrating unit 44 weights each of the second pixel values according to the shape (for example, a sphere, a polyhedron, etc.) in which the foreign matter 13 to be detected is assumed, and calculates the integrated value. By obtaining the image, the image of the foreign matter 13 in the image can be detected without omission.

The foreign matter inspection device 100 includes an image sensor 3 which is an X-ray image sensor and an image processing device 4, and the storage unit 41 stores the pixel value of the image acquired by the X-ray image sensor as the first pixel value. do.

However, the electromagnetic wave 21 is not limited to X-rays. As an example of the electromagnetic wave 21 other than X-rays, various well-known electromagnetic waves such as visible light and infrared rays can be mentioned. Along with this, it is possible to appropriately select an image sensor 3 suitable for the type of electromagnetic wave 21. A specific example of the image sensor 3 other than the X-ray image sensor is an FPD (Flat Panel Detector).

Further, as the second pixel value, the difference between the first pixel value and the background value may be used instead of the above-mentioned relative density. However, in order to estimate the volume of the foreign matter 13, it is preferable to use the relative density as the second pixel value.

Further, in the foreign matter inspection device 100, the image processing device 4 is provided outside the image sensor 3. However, the image processing device 4 may be provided inside the image sensor 3.

FIG. 9 is a diagram showing a schematic configuration of a foreign matter inspection device 101, which is a first modification of the foreign matter inspection device 100 shown in FIG. 1. Hereinafter, for convenience of explanation, the same reference numerals are given to the members having the same functions as the members described above, and the description thereof will not be repeated. The foreign matter inspection device 101 includes a moving mechanism 5 in addition to the configuration of the foreign matter inspection device 100.

It can be said that the inspection object 1 has a thickness between the side surface of the electromagnetic wave source 2 side corresponding to the incident side of the electromagnetic wave 21 and the side surface of the image sensor 3 side corresponding to the emission side of the electromagnetic wave 21. .. The direction of the thickness corresponds to the "MZ direction" in FIG. The moving mechanism 5 moves the inspection object 1 in a direction substantially perpendicular to the MZ direction. Specifically, the moving mechanism 5 moves the inspection object 1 in a certain direction along a plane substantially perpendicular to the MZ direction so as to cross between the electromagnetic wave generation source 2 and the image sensor 3. It is something that makes you. The direction parallel to the moving direction of the inspection object 1 corresponds to the "MY direction" in FIG.

Further, the image sensor 3 is a TDI (Time Delay Integration) sensor. In the foreign matter inspection device 101 in which the image sensor 3 is a TDI sensor, the foreign matter 13 adhering to the inspection object 1 is detected in the following manner.

First, while the inspection object 1 is translated by the moving mechanism 5 as described above, the electromagnetic wave 21 is applied to the inspection object 1 which is translated by the moving mechanism 5. Then, the electromagnetic wave 21 transmitted through the inspection object 1 is received by a large number of pixels 32 of the image sensor 3, and the foreign matter 13 is detected from the image of the electromagnetic wave 21 composed of the large number of pixels 32. The large number of pixels 32 form an image of the electromagnetic wave 21 at a plurality of timings, whereby an image of the same location on the inspection object 1 is formed for each inspection stage. It should be noted that each inspection stage is composed of a plurality of pixels 32 arranged in the column direction in FIG. In other words, the plurality of timings are a plurality of states in which the inspection objects 1 are located at different positions from each other. Then, by superimposing the images of the same portion on the inspection object 1 configured for each inspection stage, an image in which the foreign matter 13 is manifested is obtained, and the foreign matter 13 is detected.

According to the foreign matter inspection device 101, in the inspection performed while moving the thick inspection object 1, the deterioration of the inspection ability due to the blur is suppressed, the inspection efficiency is improved, and the risk of foreign matter detection omission occurs. Can be reduced.

FIG. 10 is a diagram showing a schematic configuration of a foreign matter inspection device 102, which is a second modification of the foreign matter inspection device 100 shown in FIG. 1. FIG. 11 is a diagram illustrating the rotation of the inspection object 1 by the rotation mechanism 6. The foreign matter inspection device 102 includes a rotation mechanism 6 in addition to the configuration of the foreign matter inspection device 100.

The rotation mechanism 6 rotates the inspection object 1 along a axis 15 extending in a direction from the incident side of the electromagnetic wave 21 toward the exit side of the electromagnetic wave 21, that is, in a direction parallel to the MZ direction. As shown in FIG. 11, the shaft 15 corresponds to the central axis of the inspection object 1.

Further, the image sensor 3 is a TDI sensor. In the foreign matter inspection device 102 in which the image sensor 3 is a TDI sensor, the foreign matter 13 adhering to the inspection object 1 is detected in the following manner.

First, the electromagnetic wave 21 is applied to the inspection object 1 rotated by the rotation mechanism 6 while the inspection object 1 is rotated by the rotation mechanism 6 as described above. The procedure from receiving the electromagnetic wave 21 transmitted through the inspection object 1 to the large number of pixels 32 of the image sensor 3 and then detecting the foreign matter 13 is the same for the foreign matter inspection device 101 and the foreign matter inspection device 102.

By performing the inspection while moving or rotating the inspection object 1 in parallel, the efficiency of the inspection can be improved. On the other hand, in an inspection performed while moving or rotating a thick inspection object 1 in parallel, there is a problem that the image of the foreign matter 13 is blurred. That is, when the foreign matter 13 adhering to the inspection object 1 is detected by using the image sensor 3 which is a TDI sensor in any of the above manners, the foreign matter 13 is blurred in the MY direction in the image configured by each inspection stage. There is. This tendency is particularly remarkable when the thickness of the inspection object 1 is large.

Since each of the foreign matter inspection device 101 and the foreign matter inspection device 102 includes an image processing device 4, particularly a pixel value integrating unit 44, the risk of omission of detection of the foreign matter 13 is reduced by the same principle as that of the foreign matter inspection device 100. can do. Similar to the foreign matter inspection device 100, the exposure time in image capture can be shortened as compared with the conventional case.

In the following, the effectiveness of the pixel value integrating unit 44 in the foreign matter inspection device 100 using X-rays as the electromagnetic wave 21 was verified. FIG. 12 is a table showing the results of the verification.

First, images of the inspection object 1 including the foreign matter 13 were taken four times and designated as "foreign matter image 1" to "foreign matter image 4" in FIG. 12, respectively. Further, the images of the inspection object 1 not including the foreign matter 13 were taken 13 times, and were designated as "foreign matter-free image 1" to "foreign matter-free image 13" in FIG. 12, respectively.

For each of these images, the density value was obtained for a certain pixel of interest 33 according to each of the methods 1 to 7 defined below. Of these, for methods 3 to 7, the integrated value of each second pixel value is used at the stage of obtaining the average value of each second pixel value.

Method 1: The third smallest value among the second pixel values of each of the nine pixels 32 provided in 3 rows and 3 columns centered on the pixel of interest 33 is set as the density value. Method 2: Centering on the pixel of interest 33 Of the second pixel values of each of the nine pixels 32 provided in 3 rows and 3 columns, the average value of the smallest value, the second smallest value, and the third smallest value is defined as the density value. Method 3: The average value of each second pixel value of a total of nine pixels 32 provided in 3 rows and 3 columns centered on the pixel of interest 33 is used as the density value. Method 4: 4 rows and 4 columns centered on the pixel of interest 33. Method 5: The average value of each second pixel value of a total of 16 pixels 32 provided in the above is used as a density value. Method 5: Each second of a total of 25 pixels 32 provided in 5 rows and 5 columns centered on the pixel of interest 33. Method 6: Using the Mean of Pixel Values as the Density Value: Each second pixel value of a total of 25 pixels 32 provided in 5 rows and 5 columns centered on the pixel 33 of interest was weighted according to a predetermined normal distribution. Method 7 using the average value of the values as the density value: Each second pixel value of a total of 25 pixels 32 provided in 5 rows and 5 columns centered on the pixel 33 of interest is different from method 6 (specifically). The mean value of the values weighted according to the normal distribution (with the dispersion of the Gaussian weight used for weighting changed) is taken as the density value. Then, for each of the methods 1 to 7, the worst S and the worst defined below. N and the worst S / worst N ratio were determined.

Worst S: Of the density values of "Foreign matter image 1" to "Foreign matter image 4", the one closest to 0% Worst N: Of the density values of "Foreign matter-free image 1" to "Foreign matter-free image 13" Worst S / Worst N ratio: Ratio of worst S to worst N If the worst S / worst N ratio is larger than 1, "Foreign matter image 1" to "Foreign matter image 4" and "Foreign matter image 4" respectively. Since it can be considered that each of "no image 1" to "no foreign object image 13" is properly distinguished, a foreign substance in a state where there is no false information (a phenomenon in which a foreign substance 13 is detected even though there is no foreign substance 13). It can be said that the detection omission of 13 is sufficiently suppressed. Further, it can be said that the larger the worst S / worst N ratio is, the clearer the distinction can be considered, and therefore the effect of suppressing the detection omission of the foreign matter 13 in the absence of false news is high.

According to FIG. 12, since the worst S / worst N ratio is larger than 1 in each of the methods 3 to 7, it can be seen that the omission of detection of the foreign matter 13 in the absence of false news is sufficiently suppressed. On the other hand, according to FIG. 12, since the worst S / worst N ratio is smaller than 1 in each of the method 1 and the method 2, it can be seen that the omission of detection of the foreign matter 13 in the absence of false news is not sufficiently suppressed. In particular, as the method 3, method 4, and method 5 move, in other words, as the number of pixels 32 for which the second pixel value is integrated increases, the worst S / worst N ratio increases, so there is no false news. It can be said that the effect of suppressing the detection omission of the foreign matter 13 in the state is enhanced.

From the above results, it was found that by integrating the second pixel values of the pixel groups belonging to the continuous region including the pixel of interest 33, the effect of suppressing the detection omission of the foreign matter 13 in the absence of false news is improved. Therefore, the effectiveness of the pixel value integrating unit 44 in the foreign matter inspection device 100 was found.

The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims, and the embodiments obtained by appropriately combining the technical means disclosed in the different embodiments. Is also included in the technical scope of the present invention.

1 Inspection target 11 Core 12 Non-aqueous electrolyte secondary battery separator 13 Foreign matter 14 Tape or label 15 Axis 2 Electromagnetic wave source 21 Electromagnetic wave 22 Center of electromagnetic wave generation part 23 Square cone 3 Image sensor 31 Main surface 32 Pixel 33 Attention Pixel 34 Reference pixel 35 Inappropriate pixel 36 to 38 Pixel group 4 Image processing device 41 Storage unit 42 Pixel value calculation unit 43 Background value setting unit 44 Pixel value integration unit 5 Movement mechanism 6 Rotation mechanism 100, 101, 102 Foreign matter inspection device

Claims (8)

In an image processing device that performs image processing on an image of electromagnetic waves transmitted through an inspection object containing foreign matter.
A storage unit that stores the first pixel value corresponding to each of a large number of pixels constituting the image, and a storage unit.
In each of the above pixels, the second pixel value of each of the above pixels is calculated, which is proportional to the difference between the first pixel value and the background value corresponding to the first pixel value after attenuation by the inspection object containing no foreign matter. Pixel value calculation unit and
A pixel value integrating unit that obtains an integrated value of each second pixel value of a pixel group belonging to a specific continuous region, and a pixel value integrating unit.
Equipped with
The integrated value is a value corresponding to the volume of the portion of the foreign matter corresponding to the pixel group, and is a value at which the volume of a part or all of the foreign matter can be estimated.
The second pixel value of each of the pixels is an image processing apparatus that is a ratio of the difference between the first pixel value and the background value of each pixel to the background value . The image processing apparatus according to claim 1 , wherein the pixel value integrating unit uses each of the large number of pixels as a pixel of interest and integrates the second pixel value of each pixel group belonging to the continuous region including the pixel of interest. The pixel value integrating unit has a second pixel value indicating a value whose attenuation is larger than a threshold value, and the second pixel value of each of a plurality of pixels forming a continuous region is integrated according to claim 1 or 2 . Image processing device. A foreign matter inspection device including the image processing device according to any one of claims 1 to 3 . In an image processing method that performs image processing on an image of electromagnetic waves that have passed through an inspection object containing foreign matter.
A storage process for storing the first pixel value corresponding to each of a large number of pixels constituting the above image, and
In each of the above pixels, the second pixel value of each of the above pixels is calculated, which is proportional to the difference between the first pixel value and the background value corresponding to the first pixel value after attenuation by the inspection object containing no foreign matter. Pixel value calculation process and
A pixel value integration process for obtaining an integrated value of each second pixel value of a pixel group belonging to a specific continuous region, and
Including
The integrated value is a value corresponding to the volume of the portion of the foreign matter corresponding to the pixel group, and is a value at which the volume of a part or all of the foreign matter can be estimated.
The second pixel value of each pixel is an image processing method in which the difference between the first pixel value and the background value in each pixel is the ratio to the background value . The image processing method according to claim 5 , wherein in the pixel value integration step, each of the large number of pixels is set as a pixel of interest, and the second pixel value of each pixel group belonging to the continuous region including the pixel of interest is integrated. In claim 5 or 6, in the pixel value integration step, claim 5 or 6 has a second pixel value indicating a value whose attenuation is larger than a threshold value, and integrates the second pixel values of each of a plurality of pixels forming a continuous region. The image processing method described. A foreign matter inspection method including the image processing method according to any one of claims 5 to 7 .

Images